strata but optimized for m5 (and maybe more later)

Round 17 blitz: mtp.metal float4 vectorization + Q8_0 gemv_catk, barrier-free simd_shuffle_down in gr_up_f/gr_up3/qsa_merge, gdn_step ib==0 readout pruning + qsa_attn char4 K loads (39.56 tok/s warm decode)


Author dawn Date Commit Parent 9ab561c3
+751 -172
4 changed files
+751 -172
4 changed files